Graphic designers are not UX designers. If that's where the author's been getting their ideas about UX, it's no wonder they don't see the value.
The ability to understand your users, then prioritize and design an experience that will serve their needs while balancing business requirements and the need for speed to get to market is far more than "common sense."
A hierarchy of things _you_ want is not going to get you to a good experience.
If you need to build an MVP quickly, and don't want your UX to suck (and you can't get a dedicated UX person on your team), I would provide this alternate set of steps:
1. Know who you're building it for. Remember the mantra "you are not your users." Even if it just means reading some published data or articles about your users, or having a chat with a few people, anything is better than nothing. Start from their perspective, not your own.
2. Tell yourself a story. The best way to design a product is to design the story of your product. Instead of starting with features and hierarchies, start by going up to the whiteboard and drawing a few user flows. Name your user, so you're thinking of them like a real person. (DON'T make complicated profiles or use silly names. It's enough to just think "would Sally want to do this?")
3. Use your story to design your features. If there was something in your flow that was vital, then use that feature. If you found yourself saying "yes, but she could always do -n- instead for now," it's non-essential, and you can leave it out.
4. Grok usability. If you don't have anyone on your team who knows what Fitt's Law is, or why the 10 Heuristics matter, find someone who does. This is where "common sense" leads most people, graphic designers especially, into pitfalls. There are a lot of UX professionals who are happy to give your product a once-over and point out the hidden traps you might have missed. You probably have a friend who does this for a living, and would love to help you.
5. Build early, iterate often. This is where the author gets it right. There's no substitute for the real world. Instead of long periods of usability testing on prototypes, if you can handle an imperfect product going live, you're better off watching how real users deal with it, and iterating quickly when problems arise. Use tools like Crazyegg, Clicktale, or Mouseflow to see what's really happening and react. Don't wait for users to tell you what's wrong, because they often can't (or won't bother to) articulate the problem for you.